CN112183528A - Method for tracking target vehicle, device, system and computer storage medium thereof - Google Patents
Method for tracking target vehicle, device, system and computer storage medium thereof Download PDFInfo
- Publication number
- CN112183528A CN112183528A CN202011007309.5A CN202011007309A CN112183528A CN 112183528 A CN112183528 A CN 112183528A CN 202011007309 A CN202011007309 A CN 202011007309A CN 112183528 A CN112183528 A CN 112183528A
- Authority
- CN
- China
- Prior art keywords
- image
- target
- picture
- vehicle body
- license plate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/63—Scene text, e.g. street names
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/62—Text, e.g. of license plates, overlay texts or captions on TV images
- G06V20/625—License plates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Abstract
The invention discloses a method, a device and a system for tracking a target vehicle and a computer storage medium, which are used for improving the identification precision of a vehicle body of the target vehicle. The target vehicle tracking method comprises the steps of analyzing a video or picture set on a road section to obtain a plurality of first license plate images and first vehicle body images; obtaining a license plate number according to the first license plate image; matching the license plate number with a preset tracking license plate number to obtain a second license plate image and a first target image; performing overlapping rate processing and azimuth angle processing on the second license plate image and all the first vehicle body images in the first target picture so as to obtain a third vehicle body image; and tracking according to the third vehicle body image and the first target picture. An apparatus, system, and computer storage medium for target vehicle tracking employs the above-described method.
Description
Technical Field
The present invention relates to the field of image recognition, and in particular, to a method, an apparatus, a system, and a computer storage medium for tracking a target vehicle.
Background
With the improvement of national life, the number of automobiles rises, and the traffic accidents are frequently caused and the hit-and-run condition is caused while convenience is brought to people. Although most road sections of the existing city are provided with the electronic eyes, the electronic eyes can analyze videos collected by the electronic eyes, judge whether a target vehicle exists or not according to license plate number information, and simultaneously acquire image information of the target vehicle so as to detect the condition of the target vehicle. However, when the traffic flow is large, a plurality of bodies may overlap, and the target vehicle body image may be erroneously recognized, and thus effective tracking may not be performed.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. Therefore, in the first aspect, the invention provides a method for tracking a target vehicle, which can improve the accuracy of vehicle body identification of the target vehicle; in a second aspect, the present invention provides an apparatus for target vehicle tracking; in a third aspect, the present invention provides a system for target vehicle tracking; in a fourth aspect, the present invention provides a computer storage medium.
According to some embodiments of the first aspect of the present invention, the method of target vehicle tracking comprises the steps of:
acquiring a video or picture set on a road section, and acquiring each picture to be detected from the video or the picture set;
performing image analysis on each picture to be detected to obtain a first license plate image set and a first vehicle body image set;
carrying out image recognition on each element in the first license plate image set in each picture to be detected to obtain a license plate number set;
matching each element in the license plate number set with a preset tracking license plate number respectively to obtain a second license plate image corresponding to the tracking license plate number from the first vehicle body image set, wherein the picture of the second license plate image is a first target picture;
respectively carrying out overlapping rate processing on the second license plate image and each element in a first vehicle body image set in the first target picture, and obtaining a plurality of second vehicle body images to be matched from the first vehicle body image set;
respectively acquiring the azimuth included angle between each second body image and the corresponding second license plate image; processing each azimuth included angle to obtain a third vehicle body image;
and tracking the target vehicle according to the third vehicle body image and the first target picture.
According to the above embodiments of the present invention, at least the following advantages are provided: according to the principle that the license plate is necessarily covered by the image of the vehicle corresponding to the license plate, and the position of the vehicle and the position of the license plate have a determined relation; and calculating a second body image through the overlapping rate. And according to the value of the azimuth included angle, a third vehicle body image is obtained, so that the vehicle body image of the target vehicle can be more accurately obtained under the condition of large traffic flow, the recognition precision of the vehicle body of the target vehicle is improved, and effective tracking is carried out.
According to some embodiments of the first aspect of the present invention, the obtaining of the orientation angle between each second body image and the corresponding second license plate image is performed separately; processing each azimuth angle to acquire a third vehicle body image, comprising the following steps:
respectively acquiring second center coordinates of the second license plate image and first center coordinates of all second body images;
acquiring a third coordinate, wherein the third coordinate is an intersection point of a y coordinate extension line of the first center coordinate and an x coordinate extension line of the second center coordinate, and the third coordinate is in one-to-one correspondence with the first center coordinate;
respectively acquiring a first connecting line of the third coordinate and the first central coordinate and a second connecting line of the second central coordinate and the first central coordinate; setting an included angle between the first connecting line and the second connecting line as an azimuth included angle, wherein the azimuth included angle is in one-to-one correspondence with the first central coordinate;
and acquiring the minimum azimuth included angle with the minimum median of all azimuth included angles, and acquiring a third vehicle body image according to the minimum azimuth included angle, wherein the third vehicle body image is a second vehicle body image corresponding to the minimum azimuth included angle.
The first center coordinates of the second body image corresponding to the target vehicle are closer to the second center coordinates than the first center coordinates of the second body images of the other vehicles. Therefore, by comparing the orientation angles of the plurality of second body images with the second license plate image to obtain the minimum orientation angle, the third body image which is most matched with the second license plate image in the plurality of second body images can be obtained.
According to some embodiments of the first aspect of the present invention, the image analyzing each picture to be detected to obtain a first license plate image set and a first vehicle body image set includes the following steps:
extracting a first license plate image of each picture to be detected;
and extracting a first vehicle body image of each picture to be detected.
By separately identifying the license plate number, the identification precision of the target vehicle can be improved.
According to some embodiments of the first aspect of the present invention, the method of target vehicle tracking further comprises the steps of:
acquiring a second target picture at the current moment;
processing the third vehicle body image at the previous moment and the second target picture based on a Kalman filtering algorithm to obtain a predicted image of the vehicle body image of the target vehicle at the current moment;
matching the predicted image of the second target picture according to a Hungarian algorithm to obtain a third vehicle body image of the target vehicle at the current moment;
and tracking the target vehicle according to the third vehicle body image and the second target picture.
Therefore, after the third vehicle body image is obtained for the first time, the predicted value of the position of the third vehicle body image at the current moment can be sequentially obtained according to the steps, so that the third vehicle body image is rapidly locked in the vicinity of the second target picture in the predicted image, and the predicted image and the surrounding area thereof are subjected to matching processing in the second picture to obtain the information of the third vehicle body image closest to the target vehicle, so that the counting amount of processing on the video or the picture set can be reduced; thereby improving the efficiency of tracking.
According to some embodiments of the first aspect of the present invention, the tracking the target vehicle according to the third body image and the second target picture comprises: and acquiring a moving track of the central coordinate of the third vehicle body image moving to the central coordinate of a third target picture, wherein the third target picture is a picture where the third vehicle body image is located. The target vehicle is always locked near the central area of the third target picture through the moving track, so that the target vehicle can be better tracked.
According to some embodiments of the first aspect of the present invention, the tracking a target vehicle according to the third body image and the second target picture further comprises: and acquiring the moving acceleration according to the central coordinate of the third vehicle body image and the size of the third target picture. By setting the moving acceleration, the moving track of the target vehicle can still be adjusted through the acceleration when the target vehicle accelerates or decelerates, and therefore the target vehicle is always in the tracking range.
According to some embodiments of the first aspect of the present invention, the obtaining of the moving acceleration according to the central coordinate of the third car body image and the size of the third target picture comprises:
acquiring a first distance from the center coordinate of the third vehicle body image to the center coordinate of the third target picture;
acquiring a second distance, wherein the second distance is half of the diagonal length of the third target picture;
and obtaining an acceleration threshold according to the first distance and the second distance, and matching the acceleration threshold with a preset acceleration value to obtain the actual acceleration.
The acceleration threshold is matched with the preset acceleration value, so that the target vehicle can be always in the central area of the third target picture.
According to some embodiments of the second aspect of the invention, the target vehicle tracking device comprises:
the data acquisition module processes the acquired video data or the image set; a storage module that stores the video data or the set of pictures; the license plate detection module acquires the video data or the picture set; outputting a license plate matching result according to the video data or the picture set; and the vehicle monitoring module outputs a third vehicle body image and a target picture according to the matching result and the video data or the picture set, and tracks a target vehicle.
All the advantages of the first aspect of the present invention are obtained because each module in the package of the target vehicle tracking device of the embodiment of the present invention performs the target vehicle tracking method according to any one of the first aspect of the present invention.
According to some embodiments of the third aspect of the invention, the target vehicle tracking system comprises:
the system comprises a data acquisition device, a data processing device and a data processing device, wherein the data acquisition device is used for acquiring video or image data on a road section; the target vehicle tracking device comprises a control module and a processing module, wherein the control module controls the flying machine body to move; the processing module executes the target vehicle tracking method according to any one of the embodiments of the first aspect to obtain a movement track; the flying machine body tracks a target vehicle according to the moving track; the remote monitoring center receives the vehicle detection result and the video or the picture of the target vehicle; and the remote monitoring center sends an instruction to the target vehicle tracking device.
Since a target vehicle tracking system of the embodiment of the present invention performs the target vehicle tracking method according to any one of the first aspect of the present invention, all the advantages of the first aspect of the present invention are obtained.
According to some embodiments of the fourth aspect of the present invention, the computer storage medium has stored thereon computer-executable instructions for causing a computer to perform the method of target vehicle tracking according to any one of the first aspect.
All the advantages of the first aspect of the present invention are obtained because the computer storage medium of the embodiment of the present invention performs the method for target vehicle tracking according to any one of the first aspect of the present invention.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a step diagram of a method of target vehicle tracking in accordance with an embodiment of the present invention;
FIG. 2 is a diagram of the tracking steps of a method of target vehicle tracking according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an azimuth angle of a target vehicle tracking method according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
In the description of the present invention, it should be understood that the orientation or positional relationship referred to in the description of the orientation, such as the upper, lower, front, rear, left, right, etc., is based on the orientation or positional relationship shown in the drawings, and is only for convenience of description and simplification of description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus, should not be construed as limiting the present invention.
In the description of the present invention, unless otherwise explicitly limited, terms such as arrangement, installation, connection and the like should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meanings of the above terms in the present invention in combination with the specific contents of the technical solutions.
It should be noted that OpenCV is a cross-platform computer vision and machine learning software library issued based on BSD license (open source), and can run on Linux, Windows, Android, and Mac OS operating systems; a number of general algorithms in image processing and computer vision are implemented; and the method is widely applied to the fields of image segmentation, face recognition, safe driving of automobiles and the like. The yolov4-tiny model is mainly used for target tracking.
Methods, apparatus, systems, and computer storage media for target vehicle tracking according to embodiments of the present invention are described below with reference to fig. 1-3.
According to some embodiments of the first aspect of the present invention, as shown in fig. 1, the method of target vehicle tracking comprises the steps of:
and S100, acquiring a video or picture set on the road section, and acquiring each picture to be detected from the video or picture set.
It should be understood that video data may be parsed into sets of pictures at different time points. In some embodiments, the video or picture set is obtained from a monitoring device on the road, such as an electronic eye, and the collected data is the picture set or video data. In other embodiments, the acquisition of the video or picture sets is derived from data from a flying device, such as a camera on a drone. In some embodiments, the video is parsed into several pictures by opencv.
Step S200, image analysis is carried out on each picture to be detected to obtain a first license plate image set and a first vehicle body image set.
It should be appreciated that in some embodiments, the first license plate image and the first body image may be identified by feeding each picture to be detected to the trained yolov4-tiny model. It should be noted that, when the first license plate image and the first body image are rectangular areas, the outlines of the license plate or the vehicle are framed and selected in the rectangular areas. It should be understood that, since the sizes of the license plate and the vehicle are substantially standard sizes, when the yolov4-tiny model is used for recognition, a threshold value of the rectangular area can be set, so that the recognized area is judged to be the first license plate image or the first vehicle body image. Since yolov4-tiny model and opencv are prior art, they are not detailed.
And S300, carrying out image recognition on each element in the first license plate image set in each picture to be detected to obtain a license plate number set.
In some embodiments, the image recognition comprises the steps of:
and carrying out license plate and background segmentation on the first license plate image so as to generate a black-and-white image.
And extracting the characters in the black-white image, and matching the characters with the standard characters to obtain the license plate number.
And S400, matching each element in the license plate number set with a preset tracking license plate number respectively to obtain a second license plate image corresponding to the tracking license plate number from the first vehicle body image set, wherein the picture of the second license plate image is a first target picture.
It should be understood that in other embodiments, the character information of the first license plate image may be extracted directly and matched to the tracking license plate number.
And S500, performing overlapping rate processing on the second license plate image and each element in the first vehicle body image set in the first target picture respectively, and obtaining a plurality of second vehicle body images to be matched from the first vehicle body image set.
It is to be understood that, according to the image recognition at S200, the coordinate set or the center coordinate of the second body image, the second license plate image in the first target picture, and the length and width of the region may be acquired. The overlapping ratio can be obtained from the coordinate calculation.
It should be noted that the overlapping ratio is a ratio of an intersection area and a union area of the first vehicle body image and the second license plate image.
In some embodiments, when the second license plate image and the first body image are acquired by the opencv method, the center coordinates of the second license plate image and the first body image and the length and width of the area may be acquired. Assuming that the center coordinates of the second card image are (x, y), the length and width of the second card image are h, w, respectively. The values of the four vertexes of the second license plate image in the x direction are x respectively1，x2(ii) a The values in the y direction are respectively y1，y2(ii) a The central coordinates of the first vehicle body image are (x ', y'), and the length and the width of the first vehicle body image are h ', w' respectively; the values of four vertexes of the first vehicle body image in the x direction are x respectively1'，x2'; the values in the y direction are respectively y1'，y2', then x1，x2,y1'，y2' represents the following:
x1＝x-w/2，x2＝x+w/2
y1＝y-h/2，y2＝y+h/2
x1'＝x'-w'/2，x2'＝x'+w'/2
y1'＝y'-h'/2，y2'＝y'+h'/2
at this time, the intersection area of the first vehicle body image and the second license plate image is xA＝max(x1,x1'),xB＝min(x2,x2'),yA＝max(y1,y1'),yb＝min(y2,y2') the region formed.
The intersection area intearea and the union _ area are as follows:
interArea＝(xB-xA)*(yB-yA)
union_area＝(x2-x1)*(y2-y1)+(x2'-x1')*(y2'-y1')-interArea
therefore, the overlap ratio IOU is equal to interArea/unity _ area.
S600, respectively obtaining the azimuth included angle between each second body image and the corresponding second license plate image; and processing each azimuth angle to acquire a third vehicle body image.
The orientation angle indicates a positional relationship between the second body image and the second signboard image.
When the traffic volume is large, the second license plate image may be covered with a plurality of second body images. The position relation between the target vehicle and the target license plate and the position relation between other vehicles and the target license plate have regularity; therefore, the third vehicle region which is most matched with the target vehicle region can be obtained according to the size of the azimuth angle.
And S700, tracking the target vehicle according to the third vehicle body image and the first target picture.
Therefore, according to the principle that the license plate is necessarily covered by the image of the vehicle corresponding to the license plate and the position of the vehicle and the license plate has a determined relation; and calculating a second body image through the overlapping rate. And according to the value of the azimuth included angle, a third vehicle body image is obtained, so that the vehicle body image of the target vehicle can be more accurately obtained under the condition of large traffic flow, the recognition precision of the vehicle body of the target vehicle is improved, and effective tracking is carried out.
In some embodiments of the first aspect of the present invention, S600, an orientation angle between each second body image and the corresponding second license plate image is respectively obtained; processing each azimuth angle to acquire a third vehicle body image, comprising the following steps:
and S610, acquiring second center coordinates of the second license plate image and acquiring first center coordinates of all second body images.
It should be understood that the first center coordinate and the second center coordinate may be obtained by establishing a two-dimensional coordinate system for the first target picture. Assume that the second center coordinate is (x)j,yj) (ii) a The first center coordinate is (x)n,yn). Where n is represented as an index of the second body image.
And S620, acquiring a third coordinate, wherein the third coordinate is an intersection point of a y coordinate extension line of the first center coordinate and an x coordinate extension line of the second center coordinate, and the third coordinate corresponds to the first center coordinate one to one.
Step S630, respectively obtaining a first connecting line of a third coordinate and the first central coordinate and a second connecting line of the second central coordinate and the first central coordinate; and setting the included angle between the first connecting line and the second connecting line as an azimuth included angle, wherein the azimuth included angle is in one-to-one correspondence with the first central coordinate.
It should be understood that, according to common knowledge, the center point of the vehicle body is necessarily higher than the center point of the license plate. The first center coordinate must therefore be in the y-axis direction of the second center coordinate. Therefore, the second body image corresponding to the first center coordinate having the y-coordinate value smaller than the second center coordinate does not satisfy the condition. And the values of the first center coordinate, the second center coordinate and the third coordinate can be determined, so that a plurality of direction included angles matched with the second body image can be obtained:
and step S640, obtaining the minimum azimuth angle with the minimum median of all azimuth angles, and obtaining a third vehicle body image according to the minimum azimuth angle, wherein the third vehicle body image is a second vehicle body image corresponding to the minimum azimuth angle.
It should be appreciated that the second body image is acquired as being closest to the target body image as the azimuth angle is smaller.
At this time, the angle between the third car body image and the second license plate image is min (β)1,β2,β3.......βn). At this time, the third vehicle body image is min (β)1,β2,β3.......βn) A corresponding second body image.
The first center coordinates of the second body image corresponding to the target vehicle are closer to the second center coordinates than the first center coordinates of the second body images of the other vehicles. Therefore, by comparing the orientation angles of the plurality of second body images with the second license plate image to obtain the minimum orientation angle, the third body image which is most matched with the second license plate image in the plurality of second body images can be obtained.
In some embodiments of the first aspect of the present invention, the step S100 of performing image analysis on each picture to be detected to obtain the first license plate image set and the first vehicle body image set includes the following steps:
and extracting a first license plate image of each picture to be detected.
And extracting a first vehicle body image of each picture to be detected.
By independently extracting the image information of the license plate and the vehicle, the accuracy of target vehicle identification can be improved.
It should be understood that when the first license plate image is extracted within the first body image, it is necessary to ensure that the first license plate image can be included when the first body image extraction is performed. However, in some existing technologies, such as yolov4-tiny, there is a problem of insufficient precision, so that the recognition frame of yolov4-tiny cannot completely frame the vehicle, and the first vehicle body image cannot contain the complete first license plate image. Thereby causing the problem that the license plate information can not be extracted.
Therefore, the accuracy of target vehicle identification can be improved by individually identifying the license plate number.
It should be noted that, in some embodiments, since the third body image can see the vehicle condition of the target vehicle, an image information formulation tracking scheme may be provided. In other embodiments, in order to realize real-time tracking, processing is further performed according to the third vehicle body image and the first target picture to ensure that the target vehicle is within the tracking range.
Thus, in further embodiments of the first aspect of the present invention, the method of target vehicle tracking further comprises the steps of:
and acquiring a second target picture at the current moment.
It should be understood that the video data or the image set need to be acquired in real time to determine the condition of the target vehicle at a certain moment of acquisition. And after the first time of third vehicle body image determination, the second target picture is the picture acquired by the acquisition equipment at the current moment in the video data or picture set in the same direction.
And processing the third vehicle body image and the second target picture at the previous moment based on a Kalman filtering algorithm to obtain a predicted image of the vehicle body image of the target vehicle at the current moment.
It should be understood that, for the detection device for tracking the target vehicle in real time, for the second target picture at different time, the position of the target vehicle at the second target picture at the corresponding time is predictable. Therefore, a prediction image can be obtained by the kalman filter algorithm, from which the image position of the vehicle body of the target vehicle in the second target picture can be roughly determined. It should be noted that, since the kalman filter algorithm is an existing target detection and tracking algorithm, a detailed description thereof is omitted here.
And performing matching processing on the predicted image of the second target picture according to the Hungarian algorithm to obtain a third car body image of the target car at the current moment.
It should be understood that, since the position of the current time is close to the position of the previous time at the adjacent and closer time, image matching of the target vehicle from the vicinity of the second coordinate set may improve the tracking efficiency.
And tracking the target vehicle according to the third vehicle body image and the second target picture.
Therefore, after the third vehicle body image is obtained for the first time, the predicted value of the position of the third vehicle body image at the current moment can be sequentially obtained according to the steps, so that the third vehicle body image is rapidly locked in the vicinity of the second target picture in the predicted image, and the predicted image and the surrounding area thereof are subjected to matching processing in the second picture to obtain the information of the third vehicle body image closest to the target vehicle, so that the counting amount of processing on the video or the picture set can be reduced; thereby improving the efficiency of tracking.
In some embodiments of the first aspect of the present invention, tracking the target vehicle according to the third body image and the second target picture comprises: and acquiring a moving track of the central coordinate of the third vehicle body image moving to the central coordinate of the third target picture, wherein the third target picture is a picture where the third vehicle body image is located. The target vehicle is always locked near the central area of the third target picture through the moving track, so that the target vehicle can be better tracked.
It should be understood that, when the target vehicle is tracked according to the third vehicle body image and the second target picture, the third target picture is the second target picture. In other embodiments, when the target vehicle is tracked according to the third vehicle body image and the first target picture, a movement track of the center coordinate of the third vehicle body image moving to the center coordinate of the third target picture is obtained. At this time, the third target picture is the first target picture.
It should be understood that, when real-time tracking is performed, the target vehicle is locked in the central area of the third target picture, so that the information of the target vehicle can be more completely exposed in the third target picture, and therefore the target vehicle can be better tracked. In some embodiments, a flying device such as an unmanned aerial vehicle is used for real-time tracking, and the flying device is controlled to move in the horizontal direction according to the distance of the moving track in the x direction.
Suppose the third target picture center coordinate is (x)center,ycenter) And the center coordinate of the third vehicle body image is (x)target,ytarget) Then the distance in the x direction is as follows:
xm＝xtarget-xcenter
wherein xmIs a distance in the x direction when x ismAnd when the value is more than 0, controlling the flying device to move leftwards, and when the value is less than 0, controlling the flying device to move rightwards until the flying device is equal to 0.
In some embodiments of the first aspect of the present invention, the second target picture tracking the target vehicle further comprises the steps of: and acquiring the moving acceleration according to the central coordinate of the third vehicle body image and the size of the third target picture. By setting the moving acceleration, the moving track of the target vehicle can still be adjusted through the acceleration when the target vehicle accelerates or decelerates, and therefore the target vehicle is always in the tracking range.
At this time, acceleration control is performed according to the distance in the y direction. The distance in the y direction is as follows:
ym＝ytarget-ycenter
wherein y ismDistance in y direction when ymWhen the speed is more than 0, the flying device is controlled to move in an accelerated way, and when the speed is less than 0, the flying device is controlled to move in a decelerated way until ymEqual to 0.
In some embodiments of the first aspect of the present invention, the obtaining of the acceleration of the movement according to the center coordinates of the third vehicle body image and the size of the third target picture comprises the steps of:
and acquiring a first distance from the center coordinate of the third vehicle body image to the center coordinate of the third target picture.
At this time, the first distance
And acquiring a second distance, wherein the second distance is half of the diagonal length of the third target picture.
At this time, the second distance is
Wherein wmThe width of the third target picture, i.e. the width in the x direction; h ismIs the height of the third target picture, i.e. the width in the y-direction.
And obtaining an acceleration threshold according to the first distance and the second distance, and matching the acceleration threshold with a preset acceleration value to obtain the actual acceleration.
It will be appreciated that using the second distance and the first distance as acceleration thresholds may control the path of movement in a centrally located area, thereby better monitoring the target vehicle. And the acceleration threshold is carried out, so that the target vehicle can be prevented from deviating due to too high or too low speed, and the third vehicle body image needs to be acquired again.
At this time, the acceleration value is assumed to be 0.8; when Distance _ m is greater than 0.8, it indicates that the target vehicle is located at a position farther from the center position, and the flying device accelerates. Otherwise, the flying device decelerates. And when the center coordinate of the third vehicle body image is near the center coordinate of the third target picture, stopping accelerating and decelerating.
In other embodiments, two acceleration values are set, which respectively represent threshold values when acceleration is needed or deceleration is needed, so as to avoid the need of re-identifying the third vehicle body image due to the fact that the speed exceeds the tracking range of the flight device too fast after acceleration.
Therefore, the acceleration threshold is matched with the preset acceleration value, so that the target vehicle can be ensured to be always in the central area of the third target picture.
According to some embodiments of the second aspect of the invention, the target vehicle tracking device comprises:
the data acquisition module processes the acquired video data or the image set; the storage module stores video data or a picture set; the license plate detection module acquires video data or a picture set; and outputting a license plate matching result according to the video data or the picture set; and the vehicle monitoring module outputs a third vehicle body image and a target picture according to the matching result and the video data or the picture set, and tracks the target vehicle.
It should be understood that, in some embodiments, the target vehicle tracking device is a real-time tracking flying device, and the data acquisition module is a camera or an infrared radar or the like arranged on the flying device for acquiring data. In other embodiments, the data acquisition module receives monitoring data transmitted from a common monitoring device, such as an electronic eye. The license plate detection module and the vehicle monitoring module both comprise image processing units, such as GPUs.
All the advantages of the first aspect of the present invention are obtained because each module in the package of the target vehicle tracking device of the embodiment of the present invention performs the target vehicle tracking method according to any one of the first aspect of the present invention.
According to some embodiments of the third aspect of the invention, a system for target vehicle tracking comprises:
the data acquisition device is used for acquiring video or image data on a road section; the target vehicle tracking device comprises a control module and a processing module, wherein the control module controls the flying machine body to move; the processing module executes the target vehicle tracking method according to any one of the embodiments of the first aspect to acquire a movement trajectory; the flying machine body tracks the target vehicle according to the moving track; the remote monitoring center receives the vehicle detection result and the video or the picture of the target vehicle; and the remote monitoring center sends an instruction to the target vehicle tracking device.
It should be understood that the flying body may be a drone or any device that can take video or picture while flying in the air.
Since a target vehicle tracking system of the embodiment of the present invention performs the target vehicle tracking method according to any one of the first aspect of the present invention, all the advantageous effects of the first aspect of the present invention are obtained.
According to some embodiments of the fourth aspect of the present invention, the computer storage medium stores computer-executable instructions for causing a computer to perform the method of target vehicle tracking according to any one of the first aspect.
All the benefits of the first aspect of the present invention are obtained in that the computer storage medium of an embodiment of the present invention performs the method of target vehicle tracking according to any one of the first aspect of the present invention.
The following describes in detail a method for performing target vehicle tracking by a target vehicle tracking apparatus according to an embodiment of the present invention with reference to fig. 1 to 3, and it is to be understood that the following description is only exemplary and not a specific limitation of the invention.
The method comprises the following steps that a target vehicle tracking device and a camera for acquiring video data are arranged on an unmanned aerial vehicle; the target vehicle tracking device executes the target vehicle tracking method. The data acquisition module is a camera; the camera is installed on unmanned aerial vehicle.
As shown in fig. 1, step S100 is to acquire a video or picture set on the road segment, and acquire each picture to be detected from the video or picture set.
Specifically, the data acquisition module acquires videos on a road section and sends the videos to the license plate detection module. The license plate detection module analyzes each frame of the video through opencv to obtain a plurality of pictures, and selects one picture from the plurality of pictures as a picture to be detected.
Step S200, image analysis is carried out on each picture to be detected to obtain a first license plate image set and a first vehicle body image set.
Specifically, each picture to be detected is transmitted to the trained yolov4-tiny model, so that a first license plate image set and a first vehicle body image set on each picture to be detected are obtained.
Specifically, the step of obtaining the first license plate image and the first vehicle body image includes the following steps:
and extracting a first license plate image of each picture to be detected.
And extracting a first vehicle body image of each picture to be detected.
Further, in step S300, image recognition is performed on each element in the first license plate image set of each picture to be detected, so as to obtain a license plate number set.
Specifically, the license plate detection module performs image recognition and executes the following steps:
and (3) carrying out license plate and background segmentation on the first license plate image so as to produce a black-and-white image.
And extracting the characters in the black-white image, and matching the characters with the standard characters to obtain the license plate number.
Further, in step S400, each element in the license plate number set is respectively matched with a preset tracking license plate number, so as to obtain a second license plate image corresponding to the tracking license plate number from the first body image set, wherein the second license plate image is located in the first target image.
Specifically, the unmanned aerial vehicle receives tracking license plate number information of the remote monitoring center and sends the tracking license plate number to the license plate detection module. And the license plate detection module compares the license plate number with the tracking license plate number. When the two images are completely consistent, the tracking is required, and the second license plate image and the picture where the second license plate image is located are output. The picture of the second license plate image is the first target picture.
At this time, the vehicle monitoring module performs the following processing:
as shown in fig. 1, step S500 is to perform overlap ratio processing on the second license plate image and each element in the first body image set in the first target picture, and obtain a plurality of second body images to be matched from the first body image set.
Specifically, two-dimensional coordinates are established in the first target picture, the central coordinates of the second license plate image are assumed to be (x, y), and the length and the width of the license plate are respectively h and w. The values of the four vertexes of the second license plate image in the x direction are x respectively1，x2(ii) a The values in the y direction are respectively y1，y2(ii) a The central coordinates of the first vehicle body image are (x ', y'), and the length and the width of the vehicle are h ', w' respectively; the values of four vertexes of the first vehicle body image in the x direction are x respectively1'，x2'; the values in the y direction are respectively y1'，y2', then x1，x2,y1'，y2' represents the following:
x1＝x-w/2，x2＝x+w/2
y1＝y-h/2，y2＝y+h/2
x1'＝x'-w'/2，x2'＝x'+w'/2
y1'＝y'-h'/2，y2'＝y'+h'/2
at this time, the intersection area of the first vehicle body image and the second license plate image is xA＝max(x1,x1'),xB＝min(x2,x2'),yA＝max(y1,y1'),yb＝min(y2,y2') the region formed.
The intersection area intearea and the union _ area are as follows:
interArea＝(xB-xA)*(yB-yA)
union_area＝(x2-x1)*(y2-y1)+(x2'-x1')*(y2'-y1')-interArea
therefore, the overlap ratio IOU is equal to interArea/unity _ area.
Further, step S600, an orientation included angle between each second body image and the corresponding second license plate image is respectively obtained; and processing each azimuth angle to acquire a third vehicle body image.
Specifically, as shown in fig. 2 and 3, a plurality of azimuth angles between all the second body images and the second license plate images are obtained; processing all the azimuth angles to acquire a third vehicle body image further comprises:
s610, acquiring second center coordinates of the second license plate image and acquiring first center coordinates of all second body images.
Specifically, assume that the second center coordinate is (x)j,yj) (ii) a The first center coordinate is (x)n,yn). Where n is represented as an index of the second body image.
Further, step S620, a third coordinate is obtained, where the third coordinate is an intersection of a y-coordinate extension line of the first center coordinate and an x-coordinate extension line of the second center coordinate, and the third coordinate corresponds to the first center coordinate one to one;
s630, respectively obtaining a first connecting line of a third coordinate and the first central coordinate and a second connecting line of a second central coordinate and the first central coordinate; and setting the included angle between the first connecting line and the second connecting line as an azimuth included angle, wherein the azimuth included angle is in one-to-one correspondence with the first central coordinate.
Specifically, as shown in fig. 3, a plurality of direction angles matched with the second body image can be obtained:
further, in step S640, a minimum azimuth angle with a minimum median value among all azimuth angles is obtained, and a third vehicle body image is obtained according to the minimum azimuth angle, where the third vehicle body image is a second vehicle body image corresponding to the minimum azimuth angle.
Specifically, the direction angle between the third car body image and the second license plate image is min (beta)1,β2,β3.......βn). At this time, the third vehicle body image is min (β)1,β2,β3.......βn) A corresponding second body image.
Further, step S700, tracking the target vehicle according to the third vehicle body image and the first target picture.
At this time, the third target picture is the first target picture.
Specifically, a movement track of the center coordinate of the third vehicle body image moving to the center coordinate of the third target picture is obtained.
Suppose the third target picture center coordinate is (x)center,ycenter) And the center coordinate of the third vehicle body image is (x)target,ytarget) Then the distance in the x direction is as follows:
xm＝xtarget-xcenter
wherein xmIs a distance in the x direction when x ismAnd when the value is more than 0, controlling the flying device to move leftwards, and when the value is less than 0, controlling the flying device to move rightwards until the flying device is equal to 0. The movement trajectory is the distance in the x direction.
Further, the moving acceleration is obtained according to the central coordinate of the third vehicle body image and the size of the third target picture.
Specifically, the vehicle monitoring module performs acceleration according to the distance in the y direction and sends a control command to the flying device to perform acceleration or deceleration. In particular, the method comprises the following steps of,assume distance y in the y-directionmThe following were used:
ym＝ytarget-ycenter
when y ismSending a control command of acceleration when the speed is more than 0, and sending a control command of deceleration when the speed is less than 0 until ymEqual to 0.
Further, the specific speed value of acceleration or deceleration sent by the vehicle monitoring module is processed as follows:
and acquiring a first distance from the center coordinate of the third vehicle body image to the center coordinate of the third target picture.
At this time, the first distance
Further, a second distance is obtained, and the second distance is half of the length of the diagonal line of the third target picture.
At this time, the second distance is
Wherein wmThe width of the third target picture, i.e. the width in the x direction; h ismIs the height of the third target picture, i.e. the width in the y-direction.
Further, an acceleration threshold value is obtained according to the first distance and the second distance, and the acceleration threshold value is matched with a preset acceleration value to obtain acceleration.
Specifically, the acceleration value is 0.8; when Distance _ m is greater than 0.8, it indicates that the target vehicle is located at a position farther from the center position, and the flying device accelerates. Otherwise, the flying device decelerates. And when the center coordinate of the third vehicle body image is near the center coordinate of the third target picture, stopping accelerating and decelerating.
After the target vehicle tracking at the current moment is finished, continuously acquiring the video and tracking the target vehicle at the current moment as follows:
and acquiring a second target picture at the current moment.
Further, the third vehicle body image and the second target picture at the previous moment are processed based on a Kalman filtering algorithm to obtain a predicted image of the vehicle body image of the target vehicle at the current moment.
Further, in the second target picture, the predicted image is subjected to matching processing based on Hungarian algorithm to obtain a third vehicle body image of the target vehicle at the current moment.
And tracking the target vehicle according to the third vehicle body image and the second target picture.
Specifically, the tracking refers to the flight trajectory and the acceleration control in step S700. At this time, the third target picture is the second target picture. And then, acquiring and processing the second target picture again at the next moment until the target vehicle does not need to be tracked or the target vehicle is not in the second target picture. When the target vehicle is not in the second target picture, the steps S100 to S600 need to be repeated until the third vehicle body image is acquired.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
The embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the gist of the present invention.
Claims (10)
1. A method of target vehicle tracking, comprising the steps of:
acquiring a video or picture set on a road section, and acquiring each picture to be detected from the video or the picture set;
performing image analysis on each picture to be detected to obtain a first license plate image set and a first vehicle body image set;
carrying out image recognition on each element in the first license plate image set in each picture to be detected to obtain a license plate number set;
matching each element in the license plate number set with a preset tracking license plate number respectively to obtain a second license plate image corresponding to the tracking license plate number from the first vehicle body image set, wherein the picture of the second license plate image is a first target picture;
respectively carrying out overlapping rate processing on the second license plate image and each element in a first vehicle body image set in the first target picture, and obtaining a plurality of second vehicle body images to be matched from the first vehicle body image set;
respectively acquiring the azimuth included angle between each second body image and the corresponding second license plate image; processing each azimuth included angle to obtain a third vehicle body image;
and tracking the target vehicle according to the third vehicle body image and the first target picture.
2. The method of target vehicle tracking according to claim 1,
respectively acquiring the azimuth included angle between each second body image and the corresponding second license plate image; the step of processing each azimuth angle to acquire a third vehicle body image comprises the following steps:
acquiring second center coordinates of the second license plate image and acquiring first center coordinates of all second body images;
acquiring a third coordinate, wherein the third coordinate is an intersection point of a y coordinate extension line of the first center coordinate and an x coordinate extension line of the second center coordinate, and the third coordinate is in one-to-one correspondence with the first center coordinate;
respectively acquiring a first connecting line of the third coordinate and the first central coordinate and a second connecting line of the second central coordinate and the first central coordinate; setting an included angle between the first connecting line and the second connecting line as an azimuth included angle, wherein the azimuth included angle is in one-to-one correspondence with the first central coordinate;
and acquiring the minimum azimuth included angle with the minimum median of all azimuth included angles, and acquiring a third vehicle body image according to the minimum azimuth included angle, wherein the third vehicle body image is a second vehicle body image corresponding to the minimum azimuth included angle.
3. The method of target vehicle tracking according to claim 1 or 2,
performing image analysis on each picture to be detected to obtain a first license plate image set and a first vehicle body image set, and the method comprises the following steps:
extracting a first license plate image of each picture to be detected;
and extracting a first vehicle body image of each picture to be detected.
4. The method of target vehicle tracking according to claim 1, further comprising the steps of:
acquiring a second target picture at the current moment;
processing the third vehicle body image at the previous moment and the second target picture based on a Kalman filtering algorithm to obtain a predicted image of the vehicle body image of the target vehicle at the current moment;
matching the predicted image of the second target picture according to a Hungarian algorithm to obtain a third vehicle body image of the target vehicle at the current moment;
and tracking the target vehicle according to the third vehicle body image and the second target picture.
5. The method of target vehicle tracking according to claim 4,
the tracking of the target vehicle according to the third vehicle body image and the second target picture comprises the following steps:
and acquiring a moving track of the central coordinate of the third vehicle body image moving to the central coordinate of a third target picture, wherein the third target picture is a picture where the third vehicle body image is located.
6. The method of target vehicle tracking according to claim 5,
the tracking of the target vehicle according to the third vehicle body image and the second target picture further comprises the following steps:
and acquiring the moving acceleration according to the central coordinate of the third vehicle body image and the size of the third target picture.
7. The method of target vehicle tracking according to claim 6,
the step of obtaining the moving acceleration according to the central coordinate of the third vehicle body image and the size of the third target picture comprises the following steps:
acquiring a first distance from the center coordinate of the third vehicle body image to the center coordinate of the third target picture;
acquiring a second distance, wherein the second distance is half of the diagonal length of the third target picture;
and obtaining an acceleration threshold according to the first distance and the second distance, and matching the acceleration threshold with a preset acceleration value to obtain the actual acceleration.
8. An apparatus for target vehicle tracking, comprising:
the data acquisition module processes the acquired video data or the image set;
a storage module that stores the video data or the set of pictures;
the license plate detection module acquires the video data or the picture set; outputting a license plate matching result according to the video data or the picture set;
and the vehicle monitoring module outputs a third vehicle body image and a target picture according to the matching result and the video data or the picture set, and tracks a target vehicle.
9. A system for target vehicle tracking, comprising:
the system comprises a data acquisition device, a data processing device and a data processing device, wherein the data acquisition device is used for acquiring video or image data on a road section;
the system comprises a flying machine body and a target vehicle tracking device arranged on the flying machine body, wherein the target vehicle tracking device comprises a control module and a processing module; the processing module executes the target vehicle tracking method according to any one of claims 1 to 7 to obtain a vehicle detection result and a movement track; the control module controls the flying machine body to move according to the moving track;
the remote monitoring center receives the vehicle detection result; and the remote monitoring center sends an instruction to the target vehicle tracking device.
10. A computer storage medium comprising, in combination,
the computer storage medium stores computer-executable instructions for causing a computer to perform the method of target vehicle tracking according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011007309.5A CN112183528B (en) | 2020-09-23 | 2020-09-23 | Method for tracking target vehicle, device, system and computer storage medium thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011007309.5A CN112183528B (en) | 2020-09-23 | 2020-09-23 | Method for tracking target vehicle, device, system and computer storage medium thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112183528A true CN112183528A (en) | 2021-01-05 |
| CN112183528B CN112183528B (en) | 2022-06-21 |
Family
ID=73957138
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011007309.5A Active CN112183528B (en) | 2020-09-23 | 2020-09-23 | Method for tracking target vehicle, device, system and computer storage medium thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112183528B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112766117A (en) * | 2021-01-10 | 2021-05-07 | 哈尔滨理工大学 | Vehicle detection and distance measurement method based on YOLOV4-tiny |
| CN113194281A (en) * | 2021-01-27 | 2021-07-30 | 广东建邦计算机软件股份有限公司 | Video analysis method and device, computer equipment and storage medium |
| CN113990101A (en) * | 2021-11-19 | 2022-01-28 | 深圳市捷顺科技实业股份有限公司 | Method, system and processing device for detecting vehicles in no-parking area |
| CN113194281B (en) * | 2021-01-27 | 2024-04-26 | 广东建邦计算机软件股份有限公司 | Video parsing method, device, computer equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080137908A1 (en) * | 2006-12-06 | 2008-06-12 | Mobileye Technologies Ltd. | Detecting and recognizing traffic signs |
| CN101877174A (en) * | 2009-09-29 | 2010-11-03 | 杭州海康威视软件有限公司 | Vehicle speed measurement method, supervisory computer and vehicle speed measurement system |
| US20160034778A1 (en) * | 2013-12-17 | 2016-02-04 | Cloud Computing Center Chinese Academy Of Sciences | Method for detecting traffic violation |
| CN106778648A (en) * | 2016-12-26 | 2017-05-31 | 江苏盛世华安智能科技股份有限公司 | Vehicle tracing and Vehicle License Plate Recognition System and recognition methods |
| CN107808392A (en) * | 2017-10-31 | 2018-03-16 | 中科信达（福建）科技发展有限公司 | The automatic method for tracking and positioning of safety check vehicle and system of open scene |
| CN111062318A (en) * | 2019-12-16 | 2020-04-24 | 桂林电子科技大学 | Sensor sharing optimal node selection method based on entropy weight method |
| CN111145555A (en) * | 2019-12-09 | 2020-05-12 | 浙江大华技术股份有限公司 | Method and device for detecting vehicle violation |
| CN111145545A (en) * | 2019-12-25 | 2020-05-12 | 西安交通大学 | Road traffic behavior unmanned aerial vehicle monitoring system and method based on deep learning |
| CN111291585A (en) * | 2018-12-06 | 2020-06-16 | 杭州海康威视数字技术股份有限公司 | Target tracking system, method and device based on GPS and dome camera |
-
2020
- 2020-09-23 CN CN202011007309.5A patent/CN112183528B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080137908A1 (en) * | 2006-12-06 | 2008-06-12 | Mobileye Technologies Ltd. | Detecting and recognizing traffic signs |
| CN101877174A (en) * | 2009-09-29 | 2010-11-03 | 杭州海康威视软件有限公司 | Vehicle speed measurement method, supervisory computer and vehicle speed measurement system |
| US20160034778A1 (en) * | 2013-12-17 | 2016-02-04 | Cloud Computing Center Chinese Academy Of Sciences | Method for detecting traffic violation |
| CN106778648A (en) * | 2016-12-26 | 2017-05-31 | 江苏盛世华安智能科技股份有限公司 | Vehicle tracing and Vehicle License Plate Recognition System and recognition methods |
| CN107808392A (en) * | 2017-10-31 | 2018-03-16 | 中科信达（福建）科技发展有限公司 | The automatic method for tracking and positioning of safety check vehicle and system of open scene |
| CN111291585A (en) * | 2018-12-06 | 2020-06-16 | 杭州海康威视数字技术股份有限公司 | Target tracking system, method and device based on GPS and dome camera |
| CN111145555A (en) * | 2019-12-09 | 2020-05-12 | 浙江大华技术股份有限公司 | Method and device for detecting vehicle violation |
| CN111062318A (en) * | 2019-12-16 | 2020-04-24 | 桂林电子科技大学 | Sensor sharing optimal node selection method based on entropy weight method |
| CN111145545A (en) * | 2019-12-25 | 2020-05-12 | 西安交通大学 | Road traffic behavior unmanned aerial vehicle monitoring system and method based on deep learning |
Non-Patent Citations (5)
| Title |
|---|
| HIROSHI NAKANO等: "Trajectory tracking control of a vehicle with a large sideslip angle", 《2015 IEEE INTERNATIONAL CONFERENCE ON ADVANCED INTELLIGENT MECHATRONICS (AIM)》 * |
| QIAN ZENG等: "A Sleep Scheduling Algorithm for Target Tracking in Energy Harvesting Sensor Networks", 《2014 INTERNATIONAL CONFERENCE ON CYBER-ENABLED DISTRIBUTED COMPUTING AND KNOWLEDGE DISCOVERY》 * |
| S. THAIPARNIT等: "Tracking Vehicles System Based on License Plate Recognition", 《2018 18TH INTERNATIONAL SYMPOSIUM ON COMMUNICATIONS AND INFORMATION TECHNOLOGIES (ISCIT)》 * |
| 刘磊等: "一种智能车载终端系统设计", 《现代电子技术》 * |
| 闫国伟: "车路协同系统中基于部件的车辆目标检测方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112766117A (en) * | 2021-01-10 | 2021-05-07 | 哈尔滨理工大学 | Vehicle detection and distance measurement method based on YOLOV4-tiny |
| CN113194281A (en) * | 2021-01-27 | 2021-07-30 | 广东建邦计算机软件股份有限公司 | Video analysis method and device, computer equipment and storage medium |
| CN113194281B (en) * | 2021-01-27 | 2024-04-26 | 广东建邦计算机软件股份有限公司 | Video parsing method, device, computer equipment and storage medium |
| CN113990101A (en) * | 2021-11-19 | 2022-01-28 | 深圳市捷顺科技实业股份有限公司 | Method, system and processing device for detecting vehicles in no-parking area |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112183528B (en) | 2022-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Malik | Fast vehicle detection with probabilistic feature grouping and its application to vehicle tracking | |
| Jazayeri et al. | Vehicle detection and tracking in car video based on motion model | |
| Cheng et al. | Lane detection with moving vehicles in the traffic scenes | |
| CN111932580A (en) | Road 3D vehicle tracking method and system based on Kalman filtering and Hungary algorithm | |
| CN112183528B (en) | Method for tracking target vehicle, device, system and computer storage medium thereof | |
| US10853963B2 (en) | Object detection device, device control system, and medium | |
| WO2005029440A1 (en) | Device for detecting road traveling lane | |
| WO2005029441A1 (en) | Device for detecting road traveling lane | |
| Muffert et al. | May I enter the roundabout? A time-to-contact computation based on stereo-vision | |
| KR102491527B1 (en) | Detection of objects in camera images | |
| US10984264B2 (en) | Detection and validation of objects from sequential images of a camera | |
| CN111213153A (en) | Target object motion state detection method, device and storage medium | |
| Liu et al. | Vehicle detection and ranging using two different focal length cameras | |
| CN110780287A (en) | Distance measurement method and distance measurement system based on monocular camera | |
| Tanaka et al. | Vehicle detection based on perspective transformation using rear-view camera | |
| Thakoor et al. | Structural signatures for passenger vehicle classification in video | |
| Fries et al. | Combining model-and template-based vehicle tracking for autonomous convoy driving | |
| Nakanishi et al. | Automatic vehicle image extraction based on spatio‐temporal image analysis | |
| CN112733678A (en) | Ranging method, ranging device, computer equipment and storage medium | |
| US20230245323A1 (en) | Object tracking device, object tracking method, and storage medium | |
| CN110727269A (en) | Vehicle control method and related product | |
| Oniga et al. | A fast ransac based approach for computing the orientation of obstacles in traffic scenes | |
| EP4177694A1 (en) | Obstacle detection device and obstacle detection method | |
| US11769337B2 (en) | Traffic signal recognition method and traffic signal recognition device | |
| Jazayeri et al. | Motion based vehicle identification in car video |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |